Legal claims defining the scope of protection, as filed with the USPTO.
1. A computer implemented method for the display of radiology, clinical, pathology, and laboratory reports as a single comprehensive medical report on a non-transitory computer-readable medium in a text, graphical or tabular format to assist in a proper diagnosis of a patient, the method comprising executing on a processor the steps of: optimizing a keyword search within a keyword searchable database of text data representing the radiology, clinical, pathology, and laboratory reports; restructuring and parsing the text data in the database to create spreadsheets corresponding to specific patient information taken from the radiology, clinical, pathology, and laboratory reports, wherein this step further comprises executing on a further processor the steps of: using a find/replace function to delete tabs, paragraphs, and colons in text; rebuilding paragraphs, or text data (or alphanumeric string), corresponding to specific patient information taken from the radiology, clinical, pathology, and laboratory reports, to create rows within a spreadsheet; reprocessing the same string of text data (or alphanumeric string) to create delimiters as placeholders for parsing the data into separate cells on the same row, each cell corresponding to one of the radiology, clinical, pathology, and laboratory reports; uploading and importing the rebuilt and processed text data (or alphanumeric string) into a spreadsheet application; directing the spreadsheet application to select the appropriate delimiters and text style; directing the spreadsheet application to select the appropriate pre-defined alphanumeric strings and text style; directing the spreadsheet application to batch replace pre-defined alphanumeric strings to allow standardization of report format; dynamically restructuring and parsing reports into text data (or alphanumeric string) (or alphanumeric strings) within cells across columns whereby the source text data (or alphanumeric string) is matched to a given row and the column content is matched as described by the column headings; dynamically enabling the “shrink to fit” feature and applying said feature to the text data (or alphanumeric string) (or alphanumeric strings) within a spreadsheet or database; dynamically using the “shrink to fit” feature when enabled as applied to text data (or alphanumeric string) (or alphanumeric string(s)) to analyze a spreadsheet or database efficiently; dynamically enabling the “fill color” feature, or background cell color, and applying said feature to all cells, and specifically only those cells, containing text data (or alphanumeric string) (or alphanumeric strings) within a spreadsheet or database; identifying cells within a spreadsheet or database by color coding to enable downstream processing or visual programming; associating cells containing text data (or alphanumeric string) (or alphanumeric strings) using said color coding; assigning a given background cell color to identify or classify the text data (or alphanumeric string) (or alphanumeric string(s)) contained within said cell; assigning a given background color to identify or classify the absence of text data (or alphanumeric string) (or alphanumeric string(s)) contained within cell; dynamically comparing the background color of a cell to another cell to generate or derive knowledge or understanding; matching of text data (or alphanumeric string) (or alphanumeric string(s)) contained within a cell, row, and/or column to a descriptive header term within a spreadsheet or database; classifying of text data (or alphanumeric string) (or alphanumeric string(s)) contained within a cell, row, and/or column to a descriptive header term within a spreadsheet or database; removing of any protected health information from said database in situations where de-identified patient data is required; and, dynamically applying the said methods to successive cells, rows, and/or columns of text data (or alphanumeric string(s)) in a spreadsheet or database; creating and applying a natural language processing engine to the database of spreadsheets for identification and translation of medical terminology contained in the spreadsheets; and applying a comprehensive automated analyzer to the natural language processed database of spreadsheets to identify and collate related medical information, wherein a result from the automated analyzer can then be viewed as the single comprehensive medical report in the text, graphical or tabular format on one of a computer or hand-held device to assist in a proper diagnosis of the patient.
2. The method according to claim 1 , wherein the step of optimizing a keyword search within the keyword searchable database further comprises the step of: using bootstrapping and jackknifing techniques on at least one of: a radiologist (physician) verified and confirmed reference standard to validate keyword search criteria for each imaging protocol; a pathologist (physician) verified and confirmed reference standard to validate keyword search criteria for each diagnostic protocol; a clinician (physician) verified and confirmed reference standard to validate keyword search criteria for each diagnosis; and, a clinician (physician) verified and confirmed reference standard to validate keyword or key-number search criteria for each laboratory protocol and parameter.
3. The method according to claim 1 , wherein the step of optimizing a keyword search within the keyword searchable database further comprises the steps of: improving the accuracy of the search results by using optimized boolean and power search techniques for queries based on at least one of: the imaging protocol in question; the diagnostic protocol in question; the diagnosis in question; the laboratory protocol or laboratory parameter in question; and repeatedly applying the search techniques to continuously improve the accuracy of the keyword search criteria to further optimize results in an effort to retrieve all records with true hits (relevant hits) while minimizing retrieval of false hits based on confounding variables.
4. The method according to claim 1 , wherein the step of optimizing a keyword search within the keyword searchable database further comprises the step of uploading the optimized results (records or reports) to a server for further processing.
5. A computer implemented method for the display of radiology, clinical, pathology, and laboratory reports as a single comprehensive medical report in a text, graphical or tabular format on a non-transitory computer-readable medium to assist in a proper diagnosis of a patient, the method comprising executing on a processor the steps of: optimizing a keyword search within a keyword searchable database of text data representing the radiology, clinical, pathology, and laboratory reports; restructuring and parsing the text data in the database into spreadsheets wherein the rows, columns and cells represent specific patient information taken from the radiology, clinical, pathology, and laboratory reports; creating and applying a natural language processing engine to the database of spreadsheets of specific patient information for identification and translation of medical terminology, wherein this step further comprises executing on a further processor the steps of: uploading a generic batch find/replace application capable of executing such task on at least one of a text file, digital document, spreadsheet, and database; applying said generic batch find/replace application to a text data (or alphanumeric string) (or alphanumeric string(s)) database; generating key text phrases, text data (or alphanumeric string) (or alphanumeric string(s)) verified as valid by a qualified professional representative to serve as a basis of a decoder table; matching the key text phrases or text data (or alphanumeric string), with or without the use of single or multi-character wildcards, to generate a decode table, decode columns, translation table, or translation columns; identifying numerical values as how far away, in single character increments within an alphanumeric string in a cell, the identified value is from a keyword (or positive AAD) to denote a level of confidence in the identified value being associated with said keyword; and, dynamically generating similar text phrases (or alphanumeric string(s)) in a repetitive manner using one of a manual input and function (formula) derivation; and, applying a comprehensive automated analyzer to the natural language processed database of spreadsheets of the patient information taken from the radiology, clinical, pathology, and laboratory reports to identify and collate related medical information, wherein a result from the automated analyzer can then be viewed as a single comprehensive medical report in the text, graphical or tabular format on one of a computer or hand-held device to assist in a proper diagnosis of the patient.
6. The method according to claim 5 , further comprising at least one of the steps of: analyzing the dynamically generated text phrases, manually or artificially derived, by a qualified professional representative to serve as the basis of a decoder table; and analyzing the dynamically generated text phrases, manually or artificially derived, by a qualified professional representative to ensure accuracy of the decoder table or translation table.
7. The method according to claim 6 , further comprising the steps of: generating at least one of a secondary color coding, a cell fill color, a background fill color, a text font or a style change to key text phrases (text data (or alphanumeric string) or alphanumeric string(s)) to serve as the basis of a decoder table; matching the at least one secondary color coding, cell fill color, background fill color, text font or style changes to the key text phrases to generate at least one of a decode table, decode columns, translation table, or translation columns; and, dynamically generating at least one of a similar secondary color coding, cell fill color, background fill color, text font or style change to key text phrases in a repetitive manner using at least one of manual input or function (formula) derivation.
8. The method according to claim 7 , further comprising the steps of: analyzing such dynamically generated secondary color coding, or cell fill color, or background fill color, and/or text font or style changes to key text phrases, manually or artificially derived, by a qualified professional representative to serve as the basis of a decoder table; and, analyzing such dynamically generated secondary color coding, or cell fill color, or background fill color, and/or text font or style changes to key text phrases, manually or artificially derived, by a qualified professional representative to ensure accuracy of the decoder table or translation table.
9. The method according to claim 8 , further comprising the steps of: dynamically applying each of the previous steps in an iterative fashion using at least one of: a different decode or translation alphanumeric string used at each step; and a binary decision making step before execution of the next step (if cell A is decoded as positive then proceed to application of decode table B to the adjacent cell), wherein the successive steps of both text data (or alphanumeric string) decoding and application of secondary color coding provides a level of natural language understanding, Quality Assurance, validation, and verification; and, dynamically changing the algorithm that decides which decoding step occurs in succession relative to the answer or result from the previous decoding step.
10. The method according to claim 9 , further comprising the steps of: dynamically decoding reports (or alphanumeric string(s)) for proper or complete performance, or performance as designed, of a diagnostic protocol based on a given reference; identifying decoded diagnostic reports as correctly performed or incorrectly performed as designed; classifying decoded diagnostic reports as correctly performed or incorrectly performed as designed; and, reclassifying incorrectly performed diagnostic reports to a correct classification.
11. The method according to claim 10 , further comprising the step of: dynamically decoding diagnostic reports, or other reports, as positive, negative, or indeterminate for a diagnosis or disease.
12. The method according to claim 10 , further comprising the step of: dynamically decoding diagnostic reports, or other reports, as positive, negative, or indeterminate, based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, for a diagnosis or disease.
13. The method according to claim 10 , further comprising the step of: dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis or disease.
14. The method according to claim 10 , further comprising at least one of the steps of: dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis or disease: based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report; or by assigning secondary color coding as generated by superimposed NLP (or automated understanding); dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, for a diagnosis or disease, by assigning secondary color coding as generated by superimposed NLP (or automated understanding); dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) by assigning secondary color coding as generated by superimposed NLP (or automated understanding), thereby representing a variably described single finding versus variably described additional finding(s), the dominant finding (or finding with the greatest number of decoded hits) can be represented by said secondary color coding; dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, thereby representing a variably described single finding versus variably described additional finding(s), the dominant finding (or finding with the greatest number of decoded hits); dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) by assigning a different secondary color coding as generated by superimposed NLP (or automated understanding), thereby representing a variably described absence/lack of a single finding versus variably described absence/lack of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded negative hits) can be represented by said secondary color coding; dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive)) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, thereby representing a variably described absence/lack of a single finding versus variably described absence/lack of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded negative hits); dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) by assigning a different secondary color coding as generated by superimposed NLP (or automated understanding), thereby representing a variably described stability (versus absence/lack) of a single finding versus variably described stability (versus absence/lack) of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded negative hits) can be represented by said secondary color coding; dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, or finding within a single cell (or examination) based on a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, thereby representing a variably described stability (versus absence/lack) of a single finding versus variably described stability (versus absence/lack) of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded negative hits); dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, finding, finding location (location within the human body), within a single cell (or examination) based on an identified numerical value (measurement or described alphanumerical size) by assigning either a predefined category (value, classification, absence or presence of disease, or other alphanumerical string) or different secondary color coding as generated by superimposed NLP (or automated understanding), thereby representing a variably described presence or stability (versus absence/lack) of a single finding versus variably described presence or stability (versus absence/lack) of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded positive or negative hits) can be represented by an assigned category (as example, value identified meets definition of aneurysmal size) or a secondary color coding; dynamically decoding diagnostic reports, or other reports, as according to known subtypes (subcategories or subclasses, both mutually exclusive and/or non-mutually exclusive) for a diagnosis (or alternative diagnosis), disease, finding, or finding location (location within the human body) within a single cell (or examination) according to a calculated mean or statistical numerical average of a given number of similarly decoded cells for a single report, based on an identified numerical value (measurement or described alphanumerical size) by assigning either a predefined category (value, classification, absence or presence of disease, or other alphanumerical string) or different secondary color coding as generated by superimposed NLP (or automated understanding), thereby representing a variably described presence or stability (versus absence/lack) of a single finding versus variably described presence or stability (versus absence/lack) of an additional finding(s), the dominant finding, or lack thereof (or finding with the greatest number of decoded positive or negative hits) can be represented by an assigned category (as example, value identified meets definition of aneurysmal size) or a secondary color coding; dynamically applying above said secondary color coding to the concept of NLP engine decoding yields thereby allowing efficient human processing or visual programming; and, dynamically applying at least one of the previous steps to successive cells, rows, and/or columns of text data (or alphanumeric string) (or alphanumeric string(s)) in a spreadsheet or database.
Unknown
July 25, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.